In an interesting case last week, a client asked if I could take a look at his company’s Google Analytics account and explain why the traffic had fallen suddenly. We discovered that it was a case of misidentified traffic, probably caused by a coding error which had just been fixed, but it illustrated how dangerous it can be if we just accept the headline numbers and not look at the data beneath.
Let’s assume that you get 100 visitors a week to your website. Or 1000. Or 10,000. It doesn’t matter. Have you looked at the various ways of segmenting them, just to see if it makes sense?
For example, in our client’s case above, if we’d looked at the traffic by country of origin, we’d have seen that most visits came from the UK, which is as you’d expect. But the next largest source, by some way, was a rather unimportant South American country. And this had suddenly stopped overnight. Not very likely, I’m sure you’ll agree, and an indication that something might be amiss. It wasn’t something any of us had thought to check. Do you?
Other things to keep an eye on include the systems people are using to visit your site (“Technology > Browser & OS”). Here you’d expect to see the main browsers such as Chrome, Safari and Internet Explorer heading the list. Then there’s the network locations (“Technology > Network”), where you’d expect to see a nice spread of different sources. Any particularly big ones should be viewed as suspect. Check for any with a lot of sessions and 100% “bounce rate”. This isn’t right.
The main report on traffic sources is also worth keeping an eye on, especially with the spambot plague we’re seeing at the moment. Again, look for ridiculously high figures in the “new sessions” or “bounce rate” columns. This will be stuff which isn’t real and which could be seriously distorting your figures. It needs to be filtered out, a job which takes time (we know, because we’re constantly doing it for nearly all of our clients!).